Can confirm, production issues ruining your design is heart breaking, especially when the issue is a purchasing department cheaping out on raw materials without consulting the engineering team,
I stuck my thermal camera in the freezer for around a minute or two, then plugged it into the phone while still in the freezer and there was no or very little gradient. I then started to use it in a room that was around 10c and image quality was improved as it was able to better distinguish the difference between cold and warm. I brought it into a room that was about 25c and image quality was very good. Before having it in the freezer, booting it up in my 10c room, Image quality was really poor and bringing it into the 25c room image quality was also less of when the camera was placed in the freezer. I think I may have an idea of what is happening, and don't think it's the hardware or manufacturing. Right now I have it in the 25c room to bring the device up to the room tempurature and I am going to see how it affects the image quality between the two rooms.
An update for the iOS app came out today, and it appears to have entirely fixed the thermal gradient issue. Mine looked just like yours did after a minute, but was OK after sitting. All appears to be well on the new software.
The white would bump the ir reflectance right up, so maybe it's related to the processor and linear reg ir emissions being reflected into the sensor during calibration?
These things are way to expensive to be playing around with. This camera has isolation issues, can't imagine how they didn't think about such a huge issue. Good thing I watched all your thermal cam breakdowns to figure it out, thanks!
Maybe the residue was solvent used with the glue. I used to have the problem as a child that when making plastic airplanes, the cockpit 'glas' became opaque when glued because of evaporating solvent. In this case as well, it seems to be in a (semi) closed of enviroment just like those cockpits...
It's now June 2019 and Seeks tech has moved on a bit...Now all models have a manual focus lens...And the pro version has a whopping 320 x 240 thermal resolution...Plus its about £100 cheaper than the equivalent Flir One.
Such a good video, even I who has no idea how these things actually work could understand what you were doing (at least somewhat). I have the same gradient issue with my unit (I'm guessing they all have?). Let's hope they can fix this through a software update.
I just got mine today after watching your previous video. I'm seeing a similar gradient of about 7 degrees F. There seems to be a semicircle in the upper left corner. Oh well, for my needs I'm in no hurry to tear this apart and trouble shoot it. Thanks for the great video.
You say you get the gradient right away from a cold start. That is different from others (including me) who have observed that the effect takes about 30 seconds to be noticeable and 2 minutes to stabilize.
Would think maybe they could put some little positioning pins on the lens assembly that'd poke through the PCB and hold in place with addition of some glue. Would mean they have to get the die placement repeatability spot on but they already place that with a machine anyway.
If this unit had the shutter outside of the lens housing, rather than between the lens and the sensor, would that allow the flat field correction to compensate for abnormalities in the lens?
The Fluke TiX1000 thermal camera would make for an interesting tear-down. At over $54,000 it might be a bit pricey though? I'd love to mount it on a quadcopter and fly it around the neighborhood at night!
I sure enjoy this stuff. I now have an E4 and am looking forward to your reflections on that. Thanks for the time, effort and fantastic video and audio! You, sir, are a master.
Very interesting. I understand the interest here is in the hardware, but I can't help but wonder if software post processing of the raw data from the board could compensate for gradient issues or other flaws resulting from manufacturing inconsistencies. In this case it seems that a gradient map could be held in software which would be applied to the live data to normalize it.
yeah, kinda. but when you go outside in winter, you would have to recalibrate as the temperature of the system changes. Although: Why not? It just requires just a surface with uniform temperature and emissivity, pressing a "manual flat field correction" button and then it would be re-calibrated. Sounds like practical solution.
The problem would be trying to mesh the flat field correction with the calibration for the lens. Since the shutter is behind the lens, the flat field correction cannot adjust for an uneven lens. If the shutter was outside the lens, however...
Yandy It doesn´t even need to be a shutter. Just a uniform piece of something. Uniform meaning in temperature and infrared emission. Thats enough hardware already. When lazy or in the fields, even a piece of paper would be good enough. Doubt about the uniformity of the emissivity or temperature of the used surface? Just take multiple pictures of a semi-uniform surface. Fixing the issue in software is brilliant - finding a semi-uniform surface in the field shouldn´t be that hard.
Hello! I've just bought seek thermal compact, but, picture from thermal camera is not clear. I mean, when I focused on a dog - it looks like a big red spot without contours. Can you help with this? I didn't find any other videos or instructions how to fix it.
Interesting. I have one and don't have any gradient issue at all. Got it about 2 weeks ago, wonder if they fixed the manufacturing issue in the meantime.
Great work Mike! I'm a little nervous now whether my unit will have the same gradient artifact. Also, I wonder why mechanical shutters are being used, surely it must be dirt cheap, easy and uniform to just mount an LCD panel in front - but I suspect it has to do with wavelengths (LCD glass would not let enough infrared light through)?
Hello from Russia, my friend, I see you are good in electronics) I have some problem with my Seek Thermal Conpact, can I get some cosultation from you? I have no answer from STC, I tried reboll and change the main CPU, the same result The consumption is about 150mA, CPU is heating, but no answer from it
smart to see the heat from the chip , bud have you got any idea how to make a microphone pre-amp powered by 48volt phantom source with a valve/transformer circuit ?
I even think that with the wire, he improved the product. "Seeing" with the camera in hand seems more intuitive to me then the screen to optical center point of offset.
Can the camera be hooked to a lcd monitor without the phone ? Is the micro usb cable coming from the camera pin combination 5 VDC and composite video out. Can this be applied to a monitor for composite video. Do you think this will work? Thank you for your help.
I need to send my CatS61 back as I'm pretty sure the charging circuit is shagged. The tested the battery and it seems fine and I've tested the test points charging points n there all over the place. I'm going to send it back to Cat as no one will work on it. Even people who work on Android phones.
Flir thermal cameras suck big time, resolution is too low, there is no support, there is no parts for it nor are there any repairers in Australia, well there is one 1,000 km away in Melbourne and they charge a flat rate to look at it of $250, may be more if it needs to be sent to USA. Needless to say i will never buy another Flir camera ever again. Warranty runs out, $600 camera dies. As for Seek, it's just as bad not worth the money so don't get fulled by repairers on here that are praising the camera's, they are getting paid to advertise it.
This video does a great service to the manufacturer. Few if any other investigators could have found the problem.
Very interesting teardown. The little sticker on the "E8" was funny.
Wow, what a cool investigation. I need to look into these things. I would love to have a small, cheap thermal imager!
How about rastering scanning a point infrared sensor? Slower, less resolution, but still neat.
spectrum.ieee.org/geek-life/hands-on/ir-eye/0
Having your design crippled due to a tiny production process error must be heartbreaking for the engineers involved in development.. =/
Sure, but this is why a good engineer designs a process, not an end result!
Can confirm, production issues ruining your design is heart breaking, especially when the issue is a purchasing department cheaping out on raw materials without consulting the engineering team,
I stuck my thermal camera in the freezer for around a minute or two, then plugged it into the phone while still in the freezer and there was no or very little gradient. I then started to use it in a room that was around 10c and image quality was improved as it was able to better distinguish the difference between cold and warm. I brought it into a room that was about 25c and image quality was very good. Before having it in the freezer, booting it up in my 10c room, Image quality was really poor and bringing it into the 25c room image quality was also less of when the camera was placed in the freezer. I think I may have an idea of what is happening, and don't think it's the hardware or manufacturing. Right now I have it in the 25c room to bring the device up to the room tempurature and I am going to see how it affects the image quality between the two rooms.
Are you nuts? Have you heard of the word "condensation" before?
An update for the iOS app came out today, and it appears to have entirely fixed the thermal gradient issue. Mine looked just like yours did after a minute, but was OK after sitting. All appears to be well on the new software.
The white would bump the ir reflectance right up, so maybe it's related to the processor and linear reg ir emissions being reflected into the sensor during calibration?
These things are way to expensive to be playing around with. This camera has isolation issues, can't imagine how they didn't think about such a huge issue. Good thing I watched all your thermal cam breakdowns to figure it out, thanks!
Maybe the residue was solvent used with the glue. I used to have the problem as a child that when making plastic airplanes, the cockpit 'glas' became opaque when glued because of evaporating solvent. In this case as well, it seems to be in a (semi) closed of enviroment just like those cockpits...
Or made in China.
Ryuuken24 The irony is that Seek manufactures their own cameras in the US (or, at least, they advertise them as such).
It's now June 2019 and Seeks tech has moved on a bit...Now all models have a manual focus lens...And the pro version has a whopping 320 x 240 thermal resolution...Plus its about £100 cheaper than the equivalent Flir One.
If the related videos are to be believed, none of this is due to dodgy electronics or poor calibration. It's proof of paranormal activity ;-)
Excellent breakdown!
Such a good video, even I who has no idea how these things actually work could understand what you were doing (at least somewhat). I have the same gradient issue with my unit (I'm guessing they all have?). Let's hope they can fix this through a software update.
I just got mine today after watching your previous video. I'm seeing a similar gradient of about 7 degrees F. There seems to be a semicircle in the upper left corner. Oh well, for my needs I'm in no hurry to tear this apart and trouble shoot it. Thanks for the great video.
Mike. Your videos are just getting better and better. Keep it up man.
Always a pleasure watching your videos.
You say you get the gradient right away from a cold start. That is different from others (including me) who have observed that the effect takes about 30 seconds to be noticeable and 2 minutes to stabilize.
Nice one im glad you figured it out for us mere mortals to understand
Ok, which is the best unit in this price range. Flir or seek? Cheers.
What about the actual lens mike? Is there room to move it in the holder, vs moving the actual holder?
Fascinating as always, Mike! Thanks so much for taking the time to post these!
Try to replace the lens with a 18mm or 25mm with focus capability. This might improve the clarity of the picture.
Just an idea.
Thanks!
Would think maybe they could put some little positioning pins on the lens assembly that'd poke through the PCB and hold in place with addition of some glue. Would mean they have to get the die placement repeatability spot on but they already place that with a machine anyway.
As always, very interesting. There are teardowns and teardowns! Thanks Mike.
If this unit had the shutter outside of the lens housing, rather than between the lens and the sensor, would that allow the flat field correction to compensate for abnormalities in the lens?
That was a very well planned job. One of the few videos on youtube that i really enjoyed and watched all the way through.
The Fluke TiX1000 thermal camera would make for an interesting tear-down. At over $54,000 it might be a bit pricey though? I'd love to mount it on a quadcopter and fly it around the neighborhood at night!
And the military drones would shoot yours down, lol
I sure enjoy this stuff. I now have an E4 and am looking forward to your reflections on that. Thanks for the time, effort and fantastic video and audio! You, sir, are a master.
Very interesting. I understand the interest here is in the hardware, but I can't help but wonder if software post processing of the raw data from the board could compensate for gradient issues or other flaws resulting from manufacturing inconsistencies. In this case it seems that a gradient map could be held in software which would be applied to the live data to normalize it.
yeah, kinda. but when you go outside in winter, you would have to recalibrate as the temperature of the system changes.
Although: Why not? It just requires just a surface with uniform temperature and emissivity, pressing a "manual flat field correction" button and then it would be re-calibrated. Sounds like practical solution.
The problem would be trying to mesh the flat field correction with the calibration for the lens. Since the shutter is behind the lens, the flat field correction cannot adjust for an uneven lens. If the shutter was outside the lens, however...
Yandy
It doesn´t even need to be a shutter. Just a uniform piece of something. Uniform meaning in temperature and infrared emission. Thats enough hardware already. When lazy or in the fields, even a piece of paper would be good enough. Doubt about the uniformity of the emissivity or temperature of the used surface? Just take multiple pictures of a semi-uniform surface.
Fixing the issue in software is brilliant - finding a semi-uniform surface in the field shouldn´t be that hard.
Hello! I've just bought seek thermal compact, but, picture from thermal camera is not clear. I mean, when I focused on a dog - it looks like a big red spot without contours. Can you help with this? I didn't find any other videos or instructions how to fix it.
Beautiful teardown!
Interesting. I have one and don't have any gradient issue at all. Got it about 2 weeks ago, wonder if they fixed the manufacturing issue in the meantime.
Mine doesn't have a gradient either. Purchased last week.
Absolutely blinding deductive skills !!! I am in awe !! Sherlock Holmes himself would be proud
Great work Mike! I'm a little nervous now whether my unit will have the same gradient artifact. Also, I wonder why mechanical shutters are being used, surely it must be dirt cheap, easy and uniform to just mount an LCD panel in front - but I suspect it has to do with wavelengths (LCD glass would not let enough infrared light through)?
Whoa, what an investigation!
Hello from Russia, my friend, I see you are good in electronics)
I have some problem with my Seek Thermal Conpact, can I get some cosultation from you?
I have no answer from STC, I tried reboll and change the main CPU, the same result
The consumption is about 150mA, CPU is heating, but no answer from it
My unit has a smaller gradient(2C) and in a different pattern.
Love your vids, would be great to see some more obscure vintage teardowns like a cheap noname cro
smart to see the heat from the chip , bud have you got any idea how to make a microphone pre-amp powered by 48volt phantom source with a valve/transformer circuit ?
Excellent sleuthing! I would have a small heart attack with damaged bond wires :)
I even think that with the wire, he improved the product.
"Seeing" with the camera in hand seems more intuitive to me then the screen to optical center point of offset.
sarowie I think he meant the little bond wires that he almost ripped with a minor collision. It would've been a dead device.
sarowie Yeah, those tiny gold bond wires...Gold is forgivingly ductile but still not the strongest thing. That would be terrifying.
i really enjoy and learn a lot from your videos. thank you very much.
have you tried hooking it to a computer?
Really excellent video as usual.
2:00 Thermal camera looks at Thermal camera. LoL )))
Can the camera be hooked to a lcd monitor without the phone ? Is the micro usb cable coming from the camera pin combination 5 VDC and composite video out. Can this be applied to a monitor for composite video. Do you think this will work?
Thank you for your help.
I don't think it will work as the app is required to view it
I highly doubt there is composite video coming out from the camera via the USB, it's much more likely a data stream similar to a webcam but not sure.
1:56 I see that Mike get an upgrade on his Flir to E8 XD
I would have put the shutter outside of the lens. Seems like a simple enough fix
Can you connect this bord to pc PC ??
Great video!
WoW, Awesome, Can you in other video tell more about circuit map and transistors
I need to send my CatS61 back as I'm pretty sure the charging circuit is shagged. The tested the battery and it seems fine and I've tested the test points charging points n there all over the place.
I'm going to send it back to Cat as no one will work on it. Even people who work on Android phones.
Awesome video, thanks for sharing!
Wonderful work Mike. So how do I fix this gradient issue?
Same way I did - snap off the lens & re-stick it in a better position.
No, I've not figured out a way to do the latter yet!
+datsthat Send it back to Seek and add a link to this, as proof of their sub par manufacturing. Ask for one without it or a refund and buy a flir one!
This was very interesting. Thanks!
Thanks!
I got a notification but this isn't in my sub box :-(
I must have one of those units!
Flir thermal cameras suck big time, resolution is too low, there is no support, there is no parts for it nor are there any repairers in Australia, well there is one 1,000 km away in Melbourne and they charge a flat rate to look at it of $250, may be more if it needs to be sent to USA. Needless to say i will never buy another Flir camera ever again. Warranty runs out, $600 camera dies. As for Seek, it's just as bad not worth the money so don't get fulled by repairers on here that are praising the camera's, they are getting paid to advertise it.